Cellular Networks Positioning Performance Analysis Reliability Part 14 pdf

Metabolic Networking through Enzymatic Sensing, Signaling and Response to Homeostatic Fluctuations 379 enables their interaction with the enzyme as shown. For instance, the depicted enzyme may catalyze the phosphorylation of glucose, whereas the breakdown products shown in the complexes (2) and (3) of Fig.1 may correspond to glyceraldehyde-3-phosphate and 3- phosphoglycerate arising on the later stages of the pathway. On the other hand, such breakdown products may interconnect different pathways within a network. For instance, 2- oxoglutarate dehydrogenase, which oxidatively decarboxylates 2-oxoglutarate to CO 2 and succinyl-CoA, may bind malonate (3C) and glyoxylic acid (2C) (Bunik & Pavlova 2006), which may be regarded as the breakdown derivatives of the 2-oxoglutarate (5C) molecule. (1) (2) (3) Fig. 1. Binding of a substrate (black in 1) and structurally relevant intermediates (black in 2 and 3), to an enzyme (grey). Certain flexibility of an enzyme structure, realized as a conformational change of its binding site in the presence of a ligand (“hand-glove” model), is schematically presented as a slight structural difference between the enzyme binding site in the free state and when ligand is bound. Competition for the enzyme binding site between the substrate and intermediate will affect enzymatic transformation of substrate. This change in the enzyme catalytic properties enables the network coordination dependent on the relative concentrations of the intermediates. In the simplest case of the competitive inhibitor, the binding of a non- transformed structural analog of the enzyme substrate blocks the correct binding of the substrate and hence blocks catalysis (Fig.1). However, taking into account multiple binding points, relatively extended structures of both the enzyme ligands and binding sites, and different enzyme-substrate complexes along the reaction pathway, multiple binding modes of ligands are possible (Fig.2), and those may have different consequences for catalysis. A good example of this type of the regulation is provided by the 2-oxoglutarate dehydrogenase, which starts the overall reaction of a key system of the tricarboxylic acid cycle, the 2-oxoglutarate dehydrogenase multienzyme complex. The cycle and its auxiliary pathways metabolize different organic acids with up to three carboxylic groups. Extensive kinetic studies of the 2-oxoglutarate dehydrogenase reaction (Bunik et al., 2000; Bunik & Pavlova, 1997a, 1997b; 1996) showed that many of these organic acids and their structural analogs, such as oxalacetate, succinate, glutarate, malonate and 2-oxomalonate, are able to bind to an enzyme-substrate complex. The partial overlap with the binding of the 2- oxoglutarate dehydrogenase substrate, 2-oxoglutarate, occurs by means of either the dicarboxylate (D) group, or both the dicarboxylate and 2-oxo (O) groups. To a certain degree which depends on the structural similarity to 2-oxoglutarate, such binding imitates the pre- catalytic complex of 2-oxoglutarate with the 2-oxoglutarate dehydrogenase (Fig.2, SE). The Cellular Networks - Positioning, Performance Analysis, Reliability 380 catalytically essential interaction with the 2-oxoglutarate dehydrogenase coenzyme, thiamine diphosphate (ThDP) is possible only when the substrate binds to the active site. In this case the formation of a pre-catalytic complex SE is followed by a conformational transition to the catalytic ES complex, where the decarboxylation of 2-oxoglutarate takes place. Following the release of the two reaction products (p 1 , p 2 ), the free enzyme E is ready for a new catalytic cycle. However, the ES complex is also able to bind another molecule, which may be a dicarboxylate (succinate, malonate, glutarate) or a 2-oxo dicarboxylate (oxalacetate), including the substrate 2-oxoglutarate itself. When bound to ES complex (Fig.2), such binding of a structural analog does not prevent catalysis. Moreover, 2- oxoglutarate may be bound to ES in either an inhibitory (S i ES) or an activatory (SES) mode. The activatory mode is associated with the slow compared to catalysis conformational Fig. 2. Enzyme regulation through the multiple binding modes of a ligand, exemplified by the interaction of 2-oxoglutarate dehydrogenase with substrate and its structural analogs. The substrate 2-oxoglutarate (S) and its dicarboxylate structural analogs with (O) or without (D) 2-oxo group may bind to the active site of the 2-oxoglutarate dehydrogenase in different modes, which depend on the enzyme state and interactions realized. The two positively charged (+) and one histidine (H) residues of the enzyme active site interact, respectively, with the two carboxyl and 2-oxo groups of a 2-oxo dicarboxylate. The precatalytic complex SE is transformed into the catalytic complex ES by the formation of the substrate adduct with the coenzyme, thiamine diphosphate (ThDP). The middle part of the figure shows the catalytic cycle. The catalytic bond cleavage occurring in the ES complex is schematically shown by dotted line. The parts below and above the catalytic cycle demonstrate the complexes formed after ES complex binds another molecule of the substrate or its analogs, correspondingly. See the text for other explanations. Metabolic Networking through Enzymatic Sensing, Signaling and Response to Homeostatic Fluctuations 381 transition of the enzyme to a more active state. The inhibitory binding S i ES is similar to that of dicarboxylates (DES) and, like the bound dicarboxylates, prevents the activation through the conformational transition. Remarkably, only the 2-oxo dicarboxylates are able to block catalysis, obviously because they more closely imitate the precatalytic complex SE, forming the three bonds essential for the SE→ES transformation (Fig.2). Binding of dicarboxylates does not prevent catalysis, but blocks the activatory transition of the enzyme due to binding the second 2-oxoglutarate molecule (complex SES). Thus, dependent on the degree of the structural similarity with the substrate, the substrate structural analogs exhibit different abilities to block catalysis as competitive inhibitors at the stage of the precatalytic complex SE, or to modify catalytic properties as regulators bound to the complex ES (Bunik & Pavlova, 1997a, 1997b). In the latter case, the binding affects the kinetically slow catalysis- associated conformational transition of the protein (Bunik et al., 1991). This phenomenon represents the enzyme hysteretic properties, also called as “enzyme memory”, being important for switching fluxes in the branch points of metabolic networks (Frieden, 1964). Thus, the network state is encoded by the concentrations of certain intermediates. An enzyme in a network may respond to the network function as a whole through dependence of catalysis on the concentrations of not only the enzyme own substrate(s) and product(s), but also some other intermediates in the network. The response may go beyond the equilibrium binding of different ligands. Indeed, it may involve the reversible, but time- dependent (slow compared to the binding and catalysis) consequences of the binding of regulatory metabolites. Affecting the slow compared to catalysis conformational transition of the enzyme, such regulation does not simply prevent the catalytic transformation, but adjusts it in the time-dependent manner, enabling the enzyme hysteretic properties. This type of sensing mechanism may thus be employed for the time-dependent regulation of distribution of the substrate-dependent fluxes through different pathways. Obviously, it can contribute to the temporal network dynamics (Kholodenko et al., 2010), which is crucial for the network function. It is remarkable that the mammalian 2-oxoglutarate dehydrogenase multienzyme complex which evolved to the very high catalytic selectivity, hardly allowing for the transformation of any natural 2-oxo acids other then 2-oxoglutarate or 2-oxoadipate, demonstrates a much less pronounced binding specificity at the level of the SE complex formation (Fig.2). That is, the enzyme does bind many of the substrate structural analogs for regulatory purposes. In this regard, it is interesting to note that the participation of the biological catalysts in the network coordination inevitably results in certain constraints to evolution of their catalytic properties to the highest binding specificity or catalytic efficiency. This may contribute to the notion that a higher level of the enzyme regulation, usually inherent in the enzymes functioning within more complex networks, is associated with a lower catalytic power compared to the less regulated orthologues (Hong et al., 1998). The view of enzymes losing catalytic power in order to satisfy regulatory requirements is also in accord with our knowledge from the directed enzyme evolution and protein engineering, which successfully create enzymes with a catalytic power higher compared to that achieved through the natural evolution. The lower catalytic power of the natural enzyme is often considered as a sub- optimal result of the evolution. However, such a view suffers from the lack of understanding that biological catalysts differ from chemical ones by the need to support not only the catalytic process per se, but also the metabolic network coordination/stability. This must impose the selection criteria additional to those increasing the catalytic efficiency and selectivity. For instance, binding of certain substrate structural analogs may interfere with Cellular Networks - Positioning, Performance Analysis, Reliability 382 the highest substrate specificity, but be required for regulation. It is also worth noting that the network coordination mechanisms and critical enzymes may differ, dependent on the network. That is, the evolution of different systems does not obligatory follow the same criteria. For instance, the high catalytic selectivity evolved in the mammalian 2-oxo acid dehydrogenase complexes, including the 2-oxoglutarate dehydrogenase one, is not an obligatory feature of these enzyme systems throughout different kingdoms of living organisms. In actinomicete Corynebacterium glutamicum a chimeric complex degrading both pyruvate and 2-oxoglutarate is present (Niebisch et al., 2006) instead of the two specific multienzyme complexes transforming either pyruvate or 2-oxoglutarate in many bacteria, plants and animals. Remarkably, the degradation of the two different substrates is performed by the two different gene products of the first component of the 2-oxo acid dehydrogenase multienzyme complex. That is, this chimeric complex uses the substrate- specific first components, like the other known 2-oxoglutarate dehydrogenase complexes. However, the overall control of the processes is changed in such a way that the oxidation of both substrates is regulated not separately, but together within one complex. The chimeric complex also has a very unusual regulation through posttranslational modification of a specific regulatory protein (Schultz et al., 2007, Niebisch et al., 2006). This example illustrates that the species-specific organization of metabolic networks and their coordination mechanisms greatly affect the catalysts evolved in living systems. Other examples illustrating this notion are also given below in Section 4. Thus, the evolution of enzymes and networks is interdependent, which creates an opportunity to change networks through changing their critical enzymes. Obviously, our understanding of the specific network coordination features and how they are met by evolutionary developments will underlie our ability to engineer cells with the designed network coordination. For instance, cells with new signaling circuits may be designed for medical and biotechnological applications (Lim, 2010). 2.1.2 Common metabolites interconnect pathways In biological networks, there are a number of substrates or coenzymes, which are common not only within a pathway, but also between different pathways. For instance, the universal energy store and donor ATP is used in many reactions of a pathway. That is, in glycolysis ATP is the energy donor in the two reactions and is synthesized from ADP in the other two reactions. Other pathways (e.g., fatty acid oxidation) also use this molecule for the same purposes. Universal substrate and coenzyme for many oxidation-reduction steps are NAD(P) + /NAD(P)H and FAD/FADH 2 , respectively. Remarkably, all these ubiquitous components of metabolic networks, as well as some important coenzymes or their derivatives, such as coenzyme A (CoA) or recently discovered thiamine adenine nucleotide (Bettendorff et al., 2007) comprise a common structural block of adenosine (adenine heterocycle bound to ribose) which connects to the catalytic part of the substrate or coenzyme (nicotinamide ring in NAD(P)H, isoalloxasine ring in FAD or thiazole ring in adenylated thiamine triphosphate) via the 5’-phosphorylated end of ribose. Moreover, special regulatory systems evolving for the coordination of more complex networks, employ regulators based on the same ubiquitous intermediates. For instance, apart from functioning as the universal redox substrate, interconnecting many pathways, nicotinamide adenine dinucleotide (NAD + ) is also the substrate for the regulatory NAD + -consuming enzymes, such as ADP-ribose transferases and poly(ADP-ribose) polymerases (Belenky et al., 2007). The former enable regulatory post-translational modifications of specific enzymes, whereas Metabolic Networking through Enzymatic Sensing, Signaling and Response to Homeostatic Fluctuations 383 poly(ADP-ribosylation) is a pluripotent cellular process important for maintenance of genomic integrity and RNA transcription in cells. Although all molecular mechanisms involved in the function of this regulatory system are not yet well characterized, the process basically depends on the depletion of the network NAD + and ATP (Du et al., 2003) and regulatory action of the intermediates accumulated upon the NAD + degradation, such as ADP-ribose (Perraud et al., 2005) and AMP (Formentini et al., 2009). Thus, the advanced regulation of the highly complex networks is based on the enzymatic sensing of the common key intermediates and their derivatives, occurring already in the primary metabolic networks. Other regulatory derivatives of nicotinamide adenine dinucleotides include the second messengers involved in calcium signaling, such as nicotinic acid-adenine dinucleotide phosphate (NAADP), which differs from NADP + by the presence of a nicotinic acid instead of a nicotinamide moiety (Rutter et al., 2008; Guse & Lee, 2008), and cyclic ADP- ribose (Graeff et al., 2009; Davis et al., 2008; Bai et al., 2005, Yue et al., 2009). As a result, common metabolites interconnecting different reactions of a pathway and different pathways of a network participate in the coordination of primary networks, whereas their derivatives regulate the metabolic networks evolved to a higher complexity. It thus appears that ubiquitous participation of these common intermediates in regulation of primary metabolic networks, associated with the wide representation of their protein binding sites, enabled evolution of the network coordination to further exploitation of these compounds in additional regulatory systems evolved. Not all of the network indicator molecules are so widely used by the network enzymes as those considered above. Instead, the indicators may also be molecules involved in the processes crucial for the network function and stability. For instance, 2–oxoglutarate is an intermediate synthesized in the tricarboxylic acid cycle. It is irreversibly degraded by either the 2-oxoglutarate dehydrogenase functioning in the cycle or prolyl hydroxylase. The former enzyme exerts an essential control on mitochondrial oxygen consumption under increasing energy demands (Bunik, 2010; Cheshchevik et al., 2010; Cooney et al., 1981), whereas the latter is the hypoxia-inducible factor which controls cellular responses to hypoxia (McDonough et al., 2006; Ginouves et al., 2008). It is therefore obvious that the decrease in the mitochondrial oxidation of 2-oxoglutarate under limited oxygen may influence availability of 2-oxoglutarate for the hypoxia-inducible factor, thus contributing to regulation of the complex network dependent on oxygen sensitivity (Ginouves et al., 2008). In good accord with this assumption, cellular sensing of the 2-oxoglutarate level is involved in retrograde signaling of mitochondria to nucleus, leading to adaptations compensating mitochondrial impairment through the changed expression of key enzymes (Butow and Avadhani, 2004; Bunik & Fernie, 2009). 2.2 Biological network buffers Some metabolites may be present in metabolic networks at very high concentrations. For instance, cells may contain 1-10 mM tripeptide glutathione (L-γ-glutamyl-L-cysteineyl- glycine) (Mieyal et al., 1995; Lopez-Mirabal & Winter, 2008) or 10-200 mM of dipeptides carnosine (beta-alanyl-L-histidine) and anserin (N2-methyl-carnosine) (Boldyrev, 2007). This creates an opportunity for the network coordination, which employs such highly abundant molecules as general intracellular sensors. For instance, the buffering or redox properties of these compounds may be used to integrate functional outcome of different enzymes which influence the pH or redox potential of intracellular milieu. The dipeptides are mostly known as general buffers and metal chelators. The latter property may be responsible for their Cellular Networks - Positioning, Performance Analysis, Reliability 384 antioxidant action, which was also suggested to be due to the direct chemical reactions between the dipeptides and reactive oxygen species (ROS) and/or the secondary products arising due to ROS, such as the products of lipid peroxidation. That is, the conjugation of carnosine with α,β-unsaturated aldehydes formed from lipid peroxidation was shown as an important mechanism for the aldehydes detoxification (Aldini et al., 2002; Guiotto et al., 2005). It should be noted, however, that in case of bifunctional lipoxidation products such reactions may also damage the protein function. For instance, the bifunctional lipoxidation- derived aldehyde 4-oxo-2-nonenal can cross-link carnosine to proteins, causing an irreversible protein modification in vitro (Zhu et al., 2009), although occurrence of such processes in vivo is not known. Significance of the very specific pattern of the dipeptide levels during development and of a number of species- and tissue-specific modifications of the basic carnosine structure is quite intriguing, but remains an enigma, as no specific protein/enzyme targets of the dipeptides have been revealed up to date (Boldyrev, 2007). Nevertheless, the exact structure of the dipeptides may be used as a chemical signature of different species, which may point to yet undefined function of such specificity for the network regulation. In contrast, the specific structure of glutathione is known to be required for its reduction by the NADPH-dependent glutathione reductase and also for the glutaredoxin-dependent deglutathionylation of proteins (Mieyal et al., 1995). Although the antioxidant function of glutathione employs its direct chemical reaction with ROS as well as that of the dipeptides, it goes beyond that, involving the bi-directional communication between the network enzymes. As will be shown in Section 2.3, the redox equilibrium between glutathione and its oxidized form, glutathione disulfide, may affect the enzymatic function of many enzymes in the network through the reversible chemical modification of the protein thiol groups. On the other hand, other enzymes affect the state of the glutathione redox buffer by generating ROS oxidizing glutathione to its disulfide and through the NADPH-dependent reduction of the glutathione disulfide (Gitler and Danon, 2003). Thus, the glutathione-dependent communication between the network enzymes is under control of the network function indicators NAD(P)H/NAD(P) + and ROS. Besides, the glutathione/glutathione disulfide redox buffer is also in redox equilibrium with other cellular thiols and disulfides of low molecular weight, which may control availability of some important SH-comprising substrates and coenzymes. Reactions between different cellular thiols and disulfides are represented by Equations 1-4, with the relative concentrations of different products dependent on the redox potentials of the participating redox couples ( * RS - / * RS-SR * and # RS - / # RS-SR # ) in the medium: * RS - + # RS-SR # ↔ * RS-SR # + # RS - (1) * RS - + * RS-SR # ↔ * RS-SR * + # RS - (2) 2 * RS - + # RS-SR # ↔ * RS-SR * + 2 # RS - (3) * RS-SR * + # RS-SR # RS − ← ⎯⎯→ 2 * RS-SR # (4) For instance, an important substrate of many enzymes, CoA, contains thiol group which enters the thiol-disulfide exchange reactions with the glutathione/glutathione disulfide (Gilbert, 1982). Furthermore, cystine or glutathione disulfide may oxidize dithiol group of reduced lipoic acid to its disulfide (Konishi et al., 1996). In living systems, lipoic acid is used as a covalently attached cofactor of the multienzyme complexes which oxidatively Metabolic Networking through Enzymatic Sensing, Signaling and Response to Homeostatic Fluctuations 385 decarboxylate 2-oxo acids (Bunik & Strumilo, 2009). In the course of catalysis, the redox active disulfide of lipoic acid undergoes redox cycling between the dithiol and disulfide forms. Addition of low molecular weight disulfides of cysteine or glutathione inhibits the catalysis in accordance with the redox potentials of the thiols and disulfides participating in the reaction, pointing to the formation of their mixed disulfides with the complex-bound lipoate (Bunik, 2000). The mixed disulfide of the complex-bound lipoic acid with glutathione was also observed in situ (Applegate et al., 2008). 2.3 Common reactivities of the enzyme functional groups and structural elements Complementary to the commonality of metabolites in the network, the enzyme structures and reactivities have much in common too. For instance, the common adenine nucleotide moiety discussed in Section 2.1.2 binds to the nucleotide binding domain of proteins, comprising an open twisted β sheet surrounded by α helices on both sides – so called Rossman fold (Branden and Tooze, 1999). The thiol/disulfide groups of the protein cysteine/cystine residues may undergo different thiol-disulfide exchange reactions (Equations 1-4) with the glutathione/glutathione disulfide or other SH/disulfide- comprising members of the network, including both the protein (thioredoxin, protein disulfide isomerase) and low molecular weight (cysteine/cystine; CoA/CoA-disulfide) components. Increasing network concentration of glutathione disulfide usually signals the prevalence of oxidative conditions and loss of the network reducing power. As a result, the thiol groups of proteins ( * RS - ) enter the reaction with glutathione disulfide ( # RS-SR # ), forming the mixed disulfide according to Equation 1 (Gilbert, 1984; Mieyal et al., 1995; Shelton et al., 2005). Similar to enzymes, the DNA-binding proteins regulating transcription may also change their function upon glutathionylation. For instance, the NFκB transcription factor is shown to be glutathionylated under hypoxic conditions in situ, with the loss of its DNA-binding activity causing the hypoxic cancer cell death (Qanungo et al., 2007). Currently, no enzyme has been shown to serve as a catalyst of this reaction (S- glutathionylation) in situ, although potential prototypes are reported, including human glutaredoxin 1 and the pi isoform of glutathione-S-transferase (Gallogly & Mieyal, 2007; Qanungo et al., 2007). The backward reaction of the protein deglutathionylation is catalyzed by glutaredoxins. The reversible post-translational modification of proteins through glutathionylation is important not only to protect the protein cysteine residues from irreversible oxidation under oxidative conditions. This modification also serves to transduce redox signals in order to either normalize homeostasis or, if this turns to be impossible, to start the death program. For instance, accumulation of hydrogen peroxide leads to depression of mitochondrial metabolism due to glutathionylation of 2-oxoglutarate dehydrogenase, with the peroxide consumption restoring both the metabolism and free thiols in the enzyme (Applegate et al., 2008). However, when the transcription factor p65- NFkappaB is S-glutathionylated, this potentiates the cell death through apoptosis (Qanungo et al., 2007). Such network switches between different states, which are dependent on the degree of homeostatic deviations, are widely spread in living systems. In particular, initial increase in the network hydrogen peroxide elicits antioxidant gene expression to reduce the peroxide, but if the peroxide level continues to increase, the prooxidant genes are expressed, inducing the cell death (Veal et al., 2007). Another important network coordination system is based on the complete oxidation- reduction of protein vicinal thiols according to Equations 1-2, where # RS-SR # depicts the Cellular Networks - Positioning, Performance Analysis, Reliability 386 protein disulfide. Usually the reductant * RS - is a small protein with the redox-active thiol/disulfide group, thioredoxin. In many enzymes this results in essential changes of the enzymatic activities, leading to metabolic switches (Gitler & Danon 2003). A classic example is the light-dark regulation of metabolic activity of chloroplasts, which is based on the light- induced reduction of the disulfide bonds in several key proteines, switching on the photosynthetic reactions of carbon fixation in plants through multiple mechanisms shown in Fig.3 (Buchanan, 1991; Danon & Mayfield, 1994; Jacquot et al., 1997; Gitler & Danon 2003). Not only metabolic enzymes, but also ion channels (Aon et al., 2007), transcription and translation factors (Chen PR, 2006; Hayashi, 1993; Jacquiersarlin & Polla, 1996; Ueno et al., 1999; Levings & Siedow, 1995; Danon & Mayfield, 1994), cytokines (Schenk et al., 1996), growth factors (Blum et al., 1996; Gasdaska et al., 1995), hormonal action (Boniface & Reichert, 1990; Makino et al., 1999) and even intercellular communication (Meng et al., 2010) may be regulated by the thioredoxin-dependent mechanism. Fig. 3. Thioredoxin-mediated regulation of chloroplastic metabolism by light. Reduction of the chloroplastic photosystem I (PS I) by light enables electron flow to the redox-protein ferredoxin (Fd) which reduced form serves as the thioredoxin (Trx) reductant in the reaction catalyzed by the ferredoxin-dependent thioredoxin reductase. Thioredoxin reduces the regulatory disulfides in several chloroplastic proteins, which leads to their activation. The latter occurs in the key enzymes of the photosynthetic pentose-phosphate pathway, the energy producing ATP synthetase and the translation-regulating protein. Many aspects of the redox regulation by reversible oxidation/reduction of protein thiols/disulfides are similar to another important posttranslational modification of proteins, involved in signal transduction, the phosphorylation/dephosphorylation (Fig.4). This modification usually occurs at the protein amino acid residues tyrosine, serine, threonin or histidine. Remarkably, the redox state of the nicotinamide adenine dinucleotides may be among the signals which are transduced by both the phosphorylation/dephosphorylation and thiol-disulfide-dependent systems. For instance, in response to increased NADH/NAD + ratio the pyruvate dehydrogenase complex is inactivated by phosphorylation (Roche et al., 2003), whereas the 2-oxoglutarate dehydrogenase complex is inactivated according to another mechanism controlled by thioredoxin (Bunik 2003). Thus, different coordination mechanisms may be involved in the transduction of the same signal through different enzymes. In general, the tight interplay between the redox states of the thiol/disulfides and nicotinamide adenine dinucleotides makes the pathways controlled by the coordination mechanism (1) in Fig.4 directly dependent on the network indicator Metabolic Networking through Enzymatic Sensing, Signaling and Response to Homeostatic Fluctuations 387 molecules (i.e. SH/-S-S-; NAD(P)H/NAD(P) + and ROS). In contrast, the phosphorylation/dephosphorylation-dependent coordination mechanism (2) does not directly depend on the cellular ADP/ATP ratio, but rather switches off/on certain pathways in response to specific signals of different nature. That is, although such signals may still be generated by depletion in cellular ATP, this would not prevent the ATP-dependent protein phosphorylation under such conditions. However, depletion of NADPH directly decreases the NADPH-dependent reduction of the glutathione disulfide. Fig. 4. Comparison of the two types of the protein regulation by post-translational modification. (1) - Modification of the protein cysteine residues in non-photosynthetic organisms; (2) – phosphorylation of the protein tyrosine, threonine, serine or histidine residues. The regulated protein is depicted as a crescent. Full oxidation/reduction of the neigbouring thiols in a protein is under the control of thioredoxin system (Trx), comprising thioredoxin and the thioredoxin reductase which in non-photosynthetic organisms is NADPH-dependent. S-glutathionylation of a protein thiol by glutathione (L-γ-glutamyl-L- cysteinyl-glycine) is controlled by glutaredoxin system (Grx), comprising glutaredoxin, glutathione and the NADPH-dependent glutathione reductase. Through catalysis by the thiol-disulfide oxidoreductases, the redox state of the nicotinamide adenine dinucleotides is related to that of cellular thiols. Both the redox states, i.e. those of the dinucleotides and thiols, are also directly related to the ROS production and scavenging. Thus, the coordination mechanism (1) may integrate signals from cellular thiols, ROS and nicotinamide adenine dinucleotides. In contrast, the coordination mechanism (2), involving the signal- and protein-specific protein kinases (PK) and protein phosphatases (PPh), does not directly depend on the overall ATP/ADP ratio. It should be noted, however, that the network evolution to increased complexity and high differentiation precludes any of the regulation types to operate at the level of whole network. Indeed, in simpler bacterial organisms certain examples exist where the protein thiol oxidation into disulfide bonds or reduction of the disulfide bonds to thiols is determined by the redox status of the environment. That is, the oxidizing milieu of the bacterial periplasm stimulates the former, whereas reducing conditions of cytoplasm SH/-S-S- NAD(P)H/NAD(P) + Trx Grx SH SH S S SH S-S-Cys γ -Glu Gly PO 3 H Signal 1 ATP ADP H 2 PO 4 PK Signal 2 PPh (1) (2) ROS Cellular Networks - Positioning, Performance Analysis, Reliability 388 promotes the latter (Eser et al., 2009). Also in mammals under some experimental settings it is possible to observe the correlation between the global intracellular change in the cellular glutathione/gluthione disulfide ratio and function of the redox-dependent transcription factors (Haddad et al., 2000). However, in view of the known reducing nature of intracellular milieu, significance of the thiol-disulfide exchange-based regulation of proteins in such a milieu had been questioned (Ziegler, 1985). For instance, the redox potential of the transcription factor OxyR is -185 mV, while thiol/disulfide redox potential of cytoplasm was estimated between -260 and -280 mV, which raises the question of how OxyR may be activated by the disulfide formation at all (Aslund et al., 1999). Further studies showed that such simplified thermodynamic consideration is not plausible for predicting the coordination mechanisms in biological networks, because the local reaction kinetics cannot be neglected, even when the overall thermodynamics is unfavorable (Danon 2002; Toledano et al., 2004). In particular, transient activation of OxyR upon local accumulation of hydrogen peroxide was supposed to occur (Aslund et al., 1999). Indeed, it was shown that activation of OxyR due to specific disulfide bond formation occurs very fast and results in a metastable protein conformation that is locally strained. The rapid kinetic reaction path and conformational strain, respectively, are supposed to drive the oxidation and reduction of OxyR (Lee et al., 2004). Also Hsp33, a member of a newly discovered heat shock protein family, is a cytoplasmically localized protein, with its reactive cysteines responding quickly to oxidizing conditions by forming the disulfide bonds. The letter activates the chaperone function of Hsp33 (Jacob et al., 1999). Thus, it was shown that the disulfide bonds can indeed be formed within the reducing intracellular environment. Moreover, the disulfide bonds can also be reduced in the oxidizing extracellular space (Hogg 2003; Yang & Loscalzo, 2005; Yang et al., 2007). Worth noting, the local accumulation of hydrogen peroxide necessary for cellular signaling through the thiol/disulfide-dependent processes is in some cases allowed by the phosphorylation-induced inactivation of peroxiredoxin I, which in its dephosphorylated state efficiently degrades hydrogen peroxide (Woo et al., 2010). This provides a good example of the cross-talk between the regulatory systems coordinating the network through the thiol-disulde exchange and phosphorylation reactions. Thus, significance of the local effects due to the thiol-disulfide exchange was established, owing to which most types of the network coordination in real systems should take into account local concentrations and kinetic effects. That is, most of the effects involving the network coordination through the posttranslational modification of the network enzymes occur transiently and in response to a local signal. 3. Coordination of the network through generation of secondary signal molecules Binding of the indicator molecules as discussed in the section 2.1 may coordinate the network through the direct regulation of a critical enzyme with the indicator bound. However, the indicator binding may also coordinate network through generation of a secondary signal molecule, which would bind to other members of the network, not binding the indicator itself. Activation of an alternative catalytic pathway in the indicator-enzyme complex may induce generation of such secondary signal. This is exemplified in, but not limited to the widely known side reaction of partial dioxygen reduction with the formation of ROS. The latter are well suited for the signaling function because their reactivity allows for chemical modification of catalysts, whereas their concentration is tightly controlled by [...]... (The picture created by Dr J Schütz) 392 Cellular Networks - Positioning, Performance Analysis, Reliability 4 Contribution of biological context to chemical networking Living systems widely use compartmentation of the medium, where metabolic networking occurs Membrane-surrounded intracellular compartments, forming different organelles, represent an advanced compartmentation mechanism, separating different... migration between the neighboring lipoate residues, the lipoate microcompartment stabilizes the thyil radical, thus promoting its interaction with the surrounding medium including the first component of the complex and/or thioredoxin 396 Cellular Networks - Positioning, Performance Analysis, Reliability Thus, the protein structure-based compartmentation may be used not only for the catalytic, but also for... that the degree of the evolutionary stabilization of such supramolecular structures depends, in particular, on the criticality of the metabolic process for the coordination of a given network and on the employed mechanisms to achieve such coordination 394 Cellular Networks - Positioning, Performance Analysis, Reliability 4.2 Advantages for the network coordination of the random coupling vs channeling... of the OxyR transcription factor by hydrogen peroxide and the cellular thiol-disulfide status Proc Natl Acad Sci U S A, Vol 96, No 11, 6161-6165 Bai, N., Lee, H.C & Laher, I (2004) Emerging role of cyclic ADP-ribose (cADPR) in smooth muscle Pharmacol Ther., Vol 105, No 2, 189-207 398 Cellular Networks - Positioning, Performance Analysis, Reliability Belenky, P., Bogan, K.L & Brenner, C (2007) NAD+... 17668-17676 400 Cellular Networks - Positioning, Performance Analysis, Reliability Gallogly, M.M & Mieyal, J.J (2007) Mechanisms of reversible protein glutathionylation in redox signaling and oxidative stress Curr Opin Pharmacol., Vol 7, No 4, 381-391 Gasdaska, J.R., Kirkpatrick, D.L., Montfort, W., Kuperus, M., Hill, S.R., Berggren, M & Powis G (1997) Oxidative inactivation of thioredoxin as a cellular growth... Shin, D.H., Kang, D., Yu, D.Y & Rhee, S.G (2010) Inactivation of peroxiredoxin I by phosphorylation allows localized H2O2 accumulation for cell signaling Cell, Vol 140 , No 4, 517-528 404 Cellular Networks - Positioning, Performance Analysis, Reliability Yang, Y & Loscalzo, J (2005) S-nitrosoprotein formation and localization in endothelial cells Proc Natl Acad Sci U S A, Vol 102, No 1, 117-122 Yang,... catalysis 4.1 Protein-based compartmentation for the network coordination Both pro- and eukaryotes employ still another compartmentation type, which does not require the membrane-separated compartments, but is created through the protein-protein interactions This type of compartmentation may be called as microcompartmentation to distinguish it from the membrane-afforded organellar compartmentation However,... Acta., Vol 1783, No 4, 629-640 Maier, T., Jenni, S & Ban, N (2006) Architecture of mammalian fatty acid synthase at 4.5 A resolution Science, Vol 311, No 5765, 1258-1262 402 Cellular Networks - Positioning, Performance Analysis, Reliability Makino, Y., Yoshikawa, N., Okamoto, K., Hirota, K., Yodoi, J., Makino, I & Tanaka, H (1999) Direct association with thioredoxin allows redox regulation of glucocorticoid... component, dihydrolipoyl acyltransferase While this route uses the energy of the 2-oxo acid oxidative decarboxylation to reduce NAD+ into NADH (reaction at the left of Fig 5), 390 Cellular Networks - Positioning, Performance Analysis, Reliability dioxygen is reduced instead of NAD+ in the side reaction (inside the circle at the right of Fig.5) As seen from the figure, the side reaction may be activated when... residues have already been included Function of this part of cellular network, which comprises many component proteins with the redox active thiols, is intensively studied, although not completely understood as yet (Thorpe & Coppock, 2007; Appenzeller-Herzog et al., 2010) However, it is well known that these reactions require separate cellular compartment where oxidizing conditions prevail to enable . J. Schütz) Cellular Networks - Positioning, Performance Analysis, Reliability 392 4. Contribution of biological context to chemical networking Living systems widely use compartmentation. binding of certain substrate structural analogs may interfere with Cellular Networks - Positioning, Performance Analysis, Reliability 382 the highest substrate specificity, but be required. intracellular milieu. The dipeptides are mostly known as general buffers and metal chelators. The latter property may be responsible for their Cellular Networks - Positioning, Performance Analysis,